[Proj] Question on colon DMS format
Gerald I. Evenden
geraldi.evenden at gmail.com
Sun May 17 09:47:17 EST 2009
On Saturday 16 May 2009 10:55:23 pm Hamish wrote:
...
> GRASS still uses this format, the following are allowed:
>
> DDDh
> (DDD.DDDDDDh is not allowed, as E has problems with
> mistranslation into e.g. e+06; so +/- DDD.DDDDD must be
> used there; could probably be coded around but currently isn't)
I do not understand the "h".
Problems with a NEWS suffix sign sounds like a library problem. On a GNU
library 11.3e is interpreted as a positive 11.3 and "e" is still in the
uninterpreded input string.
A proper lexical analyser should sense existance of an e|E followed by a
signed number. The "D" of FORTRAN infamy is not recognized. In any case
proj_dms2rad and preceeding versions have not put a restriction on ddd.dddd.
> DDD:MMh
> DDD:MM.MMMMh (recently added to deal with GPS output)
> DDD:MM:SSh
> DDD:MM:SS.SSSSSh
>
> GRASS enforces %02d for MM and SS input, but I don't think
> that's necessary and have considered removing it.
Don't use %2d as that will cause space for numbers under 10. "%d" will work
OK with proj_dms2rad.
...
> and using a degree ascii char, but then that always goes nuts
> when the LANG setting changes.
Basic, old time, standard ASCII does *not* have a degree mark and that is why
I have not used it. The degree mark is an addition to some alphabets.
I would use it but I am not a student of all the alphabets available nor do I
want to be and until a standard alphabet is mandated by the US Congress, The
United Nations, or a local star cluster federation, I must stick to old
standards although I would like to use a °.
If this august group of potential users can make an IRON-CLAD guarantee that
latin-1, UTF-8, or whatever set is univerally available to a *all* potential
users and usable in C format operations, I would add the degree mark to the
system.
Thanks for the comments.
--
The whole religious complexion of the modern world is due
to the absence from Jerusalem of a lunatic asylum.
-- Havelock Ellis (1859-1939) British psychologist
More information about the Proj
mailing list